Bioscience trends最新文献

Deep learning has brought about a revolution in the field of medical diagnosis and treatment. The use of deep learning in healthcare has grown exponentially in recent years, achieving physician-level accuracy in various diagnostic tasks and supporting applications such as electronic health records and clinical voice assistants. The emergence of medical foundation models, as a new approach to deep learning, has greatly improved the reasoning ability of machines. Characterized by large training datasets, context awareness, and multi-domain applications, medical foundation models can integrate various forms of medical data to provide user-friendly outputs based on a patien's information. Medical foundation models have the potential to integrate current diagnostic and treatment systems, providing the ability to understand multi-modal diagnostic information and real-time reasoning ability in complex surgical scenarios. Future research on foundation model-based deep learning methods will focus more on the collaboration between physicians and machines. On the one hand, developing new deep learning methods will reduce the repetitive labor of physicians and compensate for shortcomings in their diagnostic and treatment capabilities. On the other hand, physicians need to embrace new deep learning technologies, comprehend the principles and technical risks of deep learning methods, and master the procedures for integrating them into clinical practice. Ultimately, the integration of artificial intelligence analysis with human decision-making will facilitate accurate personalized medical care and enhance the efficiency of physicians.

With the development of deep learning technology, gesture recognition based on surface electromyography (EMG) signals has shown broad application prospects in various human-computer interaction fields. Most current gesture recognition technologies can achieve high recognition accuracy on a wide range of gesture actions. However, in practical applications, gesture recognition based on surface EMG signals is susceptible to interference from irrelevant gesture movements, which affects the accuracy and security of the system. Therefore, it is crucial to design an irrelevant gesture recognition method. This paper introduces the GANomaly network from the field of image anomaly detection into surface EMG-based irrelevant gesture recognition. The network has a small feature reconstruction error for target samples and a large feature reconstruction error for irrelevant samples. By comparing the relationship between the feature reconstruction error and the predefined threshold, we can determine whether the input samples are from the target category or the irrelevant category. In order to improve the performance of EMG irrelevant gesture recognition, this paper proposes a feature reconstruction network named EMG-FRNet for EMG irrelevant gesture recognition. This network is based on GANomaly and incorporates structures such as channel cropping (CC), cross-layer encoding-decoding feature fusion (CLEDFF), and SE channel attention (SE). In this paper, Ninapro DB1, Ninapro DB5 and self-collected datasets were used to verify the performance of the proposed model. The Area Under the receiver operating characteristic Curve (AUC) values of EMG-FRNet on the above three datasets were 0.940, 0.926 and 0.962, respectively. Experimental results demonstrate that the proposed model achieves the highest accuracy among related research.

The match between donor and recipient (D-R match) in the field of liver transplantation (LT) is one of the most widely debated topics today. Within the cohort of patients waiting for a transplant, better matching of the donor organ to the recipient will improve transplant outcomes, and benefit the waiting list by minimizing graft failure and the need for re-transplantation. In an era of suboptimal matches due to the sparse organ pool and the increase in extended criteria donors (ECD), ensuring adequate outcomes becomes the primary goal for clinicians in the field. The objective of this mini-review is to analyze the main variables in the evaluation of the D-R match to ensure better outcomes, the existence of scores that can help in the realization of this match, and the latest advances made thanks to the technology and development of artificial intelligence (AI).

Augmented Reality (AR) is one of the main forms of Extended Reality (XR) application in surgery. hepato-pancreato-biliary (HPB) surgeons could benefit from AR as an efficient tool for making surgical plans, providing intraoperative navigation, and enhancing surgical skills. The introduction of AR to HPB surgery is less than 30 years but brings profound influence. From the early days of projecting liver models on patients' surfaces for locating a better puncture point to today's assisting surgeons to perform live donor liver transplantation, a series of successful clinical practices have proved that AR can play a constructive role in HPB surgery and has great potential. Thus far, AR has been shown to increase efficiency and safety in surgical resection, and, at the same time, can improve oncological outcomes and reduce surgical risk. Although AR has presented admitted advantages in surgery, AR's application is still immature as an emerging technique and needs more exploration. In this paper, we reviewed the principles of AR and its developing history in HPB surgery, describing its significant practical applications over the past 30 years. Reviewing the past attempts of AR in HPB surgery could make HPB surgeons a better understanding of future surgery and the digital trends in medicine. The routine uses of AR in HPB surgery, as an indication of the operating room entering the new era, is coming soon.

Accurate ultrasound (US) image segmentation is important for disease screening, diagnosis, and prognosis assessment. However, US images typically have shadow artifacts and ambiguous boundaries that affect US segmentation. Recently, Segmenting Anything Model (SAM) from Meta AI has demonstrated remarkable potential in a wide range of applications. The purpose of this paper was to conduct an initial evaluation of the ability for SAM to segment US images, particularly in the event of shadow artifacts and ambiguous boundaries. We evaluated SAM's performance on three US datasets of different tissues, including multi-structure cardiac tissue, thyroid nodules, and the fetal head. Results indicated that SAM generally performs well with US images with clear tissue structures, but it has limited performance in the event of shadow artifacts and ambiguous boundaries. Thus, creating an improved SAM that considers the characteristics of US images is significant for automatic and accurate US segmentation.

Ultrasound image guidance is a method often used to help provide care, and it relies on accurate perception of information, and particularly tissue recognition, to guide medical procedures. It is widely used in various scenarios that are often complex. Recent breakthroughs in large models, such as ChatGPT for natural language processing and Segment Anything Model (SAM) for image segmentation, have revolutionized interaction with information. These large models exhibit a revolutionized understanding of basic information, holding promise for medicine, including the potential for universal autonomous ultrasound image guidance. The current study evaluated the performance of SAM on commonly used ultrasound images and it discusses SAM's potential contribution to an intelligent image-guided framework, with a specific focus on autonomous and universal ultrasound image guidance. Results indicate that SAM performs well in ultrasound image segmentation and has the potential to enable universal intelligent ultrasound image guidance.

The novel coronavirus disease 2019 (COVID-19) pandemic has revealed that infectious diseases will present a significant worldwide threat for a long time in the future. Centers for Disease Prevention and Control (CDCs) worldwide have developed for nearly 80 years to fight against infectious disease and protect public health. However, at the advent of the 21th century, the responsibility for prevention and control of infectious diseases has gradually been marginalized in the CDC system. The COVID-19 pandemic has also provided a glimpse into the overburdened operational process and inadequate personnel reserve of the current system of CDCs. In addition, a long-term multisectoral joint mechanism has not been created for sharing information and cooperation to facilitate public health. Reform of the system of CDCs or public health is very necessary. A global prevention and control system should be envisioned and implemented worldwide, and vertical management should be implemented throughout all levels of CDCs to improve their structure and administrative status. The WHO should expand its scope of responsibilities, especially with regard to mechanisms for joint prevention and control of infectious diseases, to substantially implement the "One Health" concept. The International Health Regulations (IHR) and relevant laws and regulations should enshrine the CDC's authority in administration and policy-making to deal with outbreaks or pandemics of infectious diseases.

Emerging infectious diseases have accompanied the development of human society while causing great harm to humans, and SARS-CoV-2 was only one in the long list of microbial threats. Many viruses have existed in their natural reservoirs for a very long time, and the spillover of viruses from natural hosts to humans via interspecies transmission serves as the main source of emerging infectious diseases. Widely existing viruses capable of utilizing human receptors to infect human cells in animals signal the possible outbreak of another viral infection in the near future. Extensive and close collaborative surveillance across nations, more effective wildlife trade legislation, and robust investment into applied and basic research will help to combat the possible pandemics of new emerging infectious diseases in the future.

In Japan, there is a growing initiative to construct centralized databases and platforms that can aggregate and manage a vast range of medical, health, and caregiving data for research and analysis. Recent advancements in artificial intelligence (AI), particularly in general-purpose models like the Segment Anything model and Chat GPT, promise significant progress towards utilizing such data-rich platforms effectively for healthcare. Traditionally, AI has displayed superior performance by learning specific images or languages, but now it is advancing towards creating models capable of learning universal traits from images and languages by training on extensive datasets. The challenge lies in the fact that these general-purpose models are trained on data that does not sufficiently incorporate medical information, making their direct application to healthcare difficult. However, the introduction of data platforms can potentially solve this problem. This would lead to the development of universally applicable models to process medical images and AI assistants that can support both doctors and patients. These medical AI assistants can function as a "sub-doctor" with extensive knowledge, assisting in comprehensive analysis of symptoms, early detection of rare diseases, and more. They can also serve as an intermediary between the doctor and the patient, helping to simplify consultations and enhance patient understanding of medical conditions and treatments. By bridging this gap, the AI assistant can help to reduce doctors' workload, improve the quality of healthcare, and facilitate early detection and prevention in the elderly population.

Detecting and appropriately diagnosing a Mycobacterium tuberculosis infection remains technologically difficult because the pathogen commonly hides in macrophages in a dormant state. Described here is novel near-infrared aggregation-induced-emission luminogen (AIEgen) labeling developed by the current authors' laboratory for point-of-care (POC) diagnosis of an M. tuberculosis infection. The selectivity of AIEgen labeling, the labeling of intracellular M. tuberculosis by AIEgen, and the labeling of M. tuberculosis in sputum samples by AIEgen, along with its accuracy, sensitivity, and specificity, were preliminarily evaluated. Results indicated that this near-infrared AIEgen labeling had satisfactory selectivity and it labeled intracellular M. tuberculosis and M. tuberculosis in sputum samples. It had a satisfactory accuracy (95.7%), sensitivity (95.5%), and specificity (100%) for diagnosis of an M. tuberculosis infection in sputum samples. The current results indicated that near-infrared AIEgen labeling might be a promising novel diagnostic tool for POC diagnosis of M. tuberculosis infection, though further rigorous verification of these findings is required.